SE Fixed Bias Pentode Amps

I decided to build a new amp tonight that can run 6V6, 6F6, etc, so it needs to be fixed bias. But I ran into an odd little gap in my knowledge here. You got the bias voltage going to the grid, and the cathode has nothing, do you just ground it? This all seems really bizzare to me, I just can not part with the idea of a resistor being on the cathode.

adam

Put a 1 ohm resistor between the cathode and ground and bring each end to test point jacks on the chassis. Now you can check the idle current in milliamps (yes, the screen current is included) by reading millivolts on your DMM. Makes for easy bias checking/setting. I got this idea from Doug Hoffman's amp site, but I'm sure others have done this as well.

But to answer your question, yes, with a bias voltage applied to the grid, you can just ground the cathode.

Yeah, I just went ahead and did that last night, and it is working nicely. But I will need to add that 1 Ohm resistor, I do not know why I did not think of that last night, I have added these resistors to quite a few amps. Fixed bias SE just seems so weird to me, but it is nice to be able to just swap different kinds of power tubes and rebias nice and easy.

What about caps on the cathode in fixed bias, do they have any use of effect?

thanks,
adam

Hmm. Does stuff like the UniValve pull any fancy tricks to adjust the bias? I thought that cathode bias was generally pretty forgiving for tube type swaps (allowing that everything else is happy, like filiment supply and OT headroom).

Bear

> can run 6V6, 6F6, etc, so it needs to be fixed bias.

Doesn't "need" to be fixed (actually user-adjusted) bias.

Stay with the classic tubes. Don't go too close to the voltage and current ratings of the smallest ones. That means using transformer impedance suited for the smallest of the set. Check pinouts, there are some small differences.

If you use 250V on the plate, a bit less on the screen, 8K-10K load, and a 470Ω cathode resistor, and check pinouts, you can swap-in any of 6F6, 6K6, 6V6/6AQ5, 6EY6, 6EZ6, 6L6, 7027 and even 7591 and 6550 without re-biasing. All of these will wind up at about the maximum power point for 250V B+ and 8K-10K load.

Of course it is a little silly to take 3 Watts from a mighty 6L6GC or 7027. But to get more power you not only have to change the cathode resistor, you have to change the load impedance and often the B+ as well. 6L6 will make a lot more power than 6F6 IF you drop the load impedance to 2K-3K (and re-bias for more current), and you can also (for 6L6GC) raise the plate voltage a lot.

What you will really find is that all tubes are nearly the same when worked at the same B+ and load, and an appropriate current. There are three main classes: US pentode like 6F6, the 6L6, and the euro KT pentodes. A 6V6 is very much like half a 6L6, and if you run a 6L6 at 6V6 currents and loads you may not be able to tell them apart. A 6550 is not so different from two 6L6. In single-ended, the 6F6 will make less even-order distortion than the 6L6/6V6 family. Most of the KT series will have less abrupt clipping than the 6L6 family. But if you design an amp to take "any tube", you have to design for the smallest of the set, in which case the bigger members of each family sound and bias almost exactly like their smaller kin, just eating more heater power to support the current you are not asking them to give.

So is the effect limited, then, to kind of a cork-sniffing, fine distinction level? Is there some self-adjusting cirucitry in something like the THD UniValve that changes things up when output tubes are swapped?

I suppose one way to go about this is to be able to switch in and out some extra resistance in the supply to be able to drop some voltage (or raise it, depending on your perspective). I had contemplated doing this with the rather inefficient method of tube rectifier swaps as a means of selecting the voltage drop, but SE is not supposed to get much rectifier sag effect, so diodes or FRED's make sense as a matter of economy.

(Is it resistance or capacitance that folks are complaining about when they talk about "overfiltering" a guitar amp power supply?)

(Afterthought - there is the Hammond SE experimenter type iron where you can re-hook it to whatever taps for the appropriate output load. I know some people think they are a bit audiophile and not very rock and roll, but it's a way at the tweakability problem.)

Bear

(Wow, is this a good idea or plain stupid: In addition to the speaker impedence tap switching on the backside of the OT, what about tap switching on the front side with iron like the Hammond?)

Bear

One of the good things about using a cathode resistor for bias is that it is kind of self compensating, so you can plug in different tube types and it will sort of set them up in the right area. As the plate curent increases, the voltage across the cahode resistor increases, which means the grid becomes more negative, which causes the plate current to decreas to the balanced point.

So if you plug a 6L6 into a 6V6 spot, the increase in current will increase the bias automatically.
Thus, the term "self bias".

So you are better off with self bias than fixed bias if you don't want to have to turn a bias pot all the time.

> what about tap switching on the front side...is this a good idea or plain stupid...

Yes. (a little of both...)

This is actually a kewl way to get various "Soundz" out of a hard-driven power pentode.

Basically an equivalent of monkeying with which impedence tap you have on the same load of speaker, then?

Bear

This amp I am makeing is just one out of boredom useing what I have at hand and built into the chassis of an old Organ extension cab that some one else tried to rewire and failed at it. Just useing parts I have sitting around. Which is another reason I went with fixed bias, did not have any suitable resistors sitting about for the cathode.

I chose to use the fixed bias since I am useing a Hammond 125ESE output transformer and can switch the impedance when I change tubes. I already have a 6V6 amp that is cathode biased, and it is quite tolerant of running other kinds of output tubes, but you do not get nearly as much variation in sound when switching tubes as you get with the 125ESE and fixed bias.

(Wow, is this a good idea or plain stupid: In addition to the speaker impedence tap switching on the backside of the OT, what about tap switching on the front side with iron like the Hammond?)

Click to expand...

The Hammond SE transformers do not have multipul primaries, just multipul secondaries.

Is there some self-adjusting cirucitry in something like the THD UniValve that changes things up when output tubes are swapped?

Click to expand...

The THD amps use some sort of solid state current sensing circuitry, There is a webpage somewhere That gives a pretty good explanation on whats going on in these amps, and they have patents on it I am sure so the patant archives should be able to tell you alot.

Afterthought - there is the Hammond SE experimenter type iron where you can re-hook it to whatever taps for the appropriate output load.

Click to expand...

That is the 125 experimenter series, I am useing the 125ESE, which is the 15 watt version. They are great transformers, very versitile and sound excellant. Even some of the audiophiles really like them. Not to mention the $35 price, best output transformer I have found for SE guitar amps. Large air gapped core and very under rated as far as specs go, I have had jsut over 20 watts through one of these and it took it very well, not even a hint of saturation.

adam

[quote author="adamasd"]
The Hammond SE transformers do not have multipul primaries, just multipul secondaries.

[/quote]

Ah, I finally bothered to check out the wiring schemes. Tricky buggers. I saw the Angela Electronics listing where they listed three primary resistances and just assumed.

Bear

I have the first glass audio. That issue has a servo bias of ouput tubes.

I first read about solid state servo bias of tubes back around 1977. I high end hifi power amp that used color TV sweep tubes had solid state control of output tube operating points.

I think it used solid state cmos logic ands opamps IIRC. I can not remember right now what brand it was I seem to remember the designer worked at the NST(now NIST) Talk about ahead of the time design. I think the company name started with B. I might have the ad brochure stored somewhere.

If the amp uses servo control of bias that is an old idea. I have built guitar overdrives with servo control years ago.

It occurs to me that there is a "new" way to self-bias power tubes.

Maybe.

They have these new "remote temperature sensors". There is one you stick in your ear (your child's ear) and get a body-temp reading from down around the ear-drum. There is one for mechanics that has a telescope and can read spot temps from several feet away. Good for checking exhaust manifolds, or industrial bearings.

Tube dissipation is limited by plate and seal temperature. The seals are designed so if the plate is OK, the seals will be OK (ignoring special cases like transmitting tubes that must have a fan on their base seals). The max temp of the plate varies with how much you paid: there is good and gooder plate-stuff. But in our tubes, not a big difference. Maybe (making up numbers) 600 for a cheap 6V6, 800 for a Tungsol 6550. The difference in plate rating, 12W or 35W, is mostly about plate size, not so much plate temperature.

So if you could read the plate temperature with an IR sensor: start with heavy bias (tube cut off), slowly pass more current. When the temperature comes up to 600 or so, set the bias there. That puts you at nearly rated dissipation, which is often the best for sound and power in single-ended tube amps.

Problem: can't work on metal-case 6L6, probably not on black-coated 6V6, and may not work at all. Glass looks clear to the eye but is actually opaque to IR heat rays. So these IR-optical sensors may only measure the glass temp, not the plate temp. And there are much easier ways to measure glass-temp. And I suspect that limiting glass temperature varies much more than limiting plate temp.

That's a clever idea PRR. 'Tis true that the transmission of glass varies a lot with wavelength and type of glass. Pursuing the idea further one could do a two- or more band measurement and fold in the Planck curve for blackbody radiation (most plates will turn out to be fairly black I believe). Then one could get the temperature of the plate disentangled from that of the envelope methinks.

Brad

If you can manage to seperate the envelope and plate in your temp readings, wouldn't you still run into problems with old tubes that have alot of distortion in the glass. Or if you threw in a tube with an ST shaped envelope. Or would refreaction caused by the glass not be an issue at these wavelengths?

adam